Glossary
Updated
A governance layer that treats each AI deployment as a distinct “use case,” requiring a standardized intake form capturing objectives, data sources, impact assessments, and risk ratings. Approved use cases are cataloged, periodically reviewed for ongoing compliance, and sunset when objectives change. Governance ensures that new or modified use cases undergo impact and risk assessments before any production rollout.
Example
A healthcare provider’s AI governance office requires every proposed use case (e.g., readmission prediction, image analysis) to complete a use-case template. They review each template to confirm data privacy, bias mitigation plans, and risk-level approvals. The office tracks each use case’s status quarterly, retiring those no longer in scope.